Managing Typhoon Risk in China

Home , Typhoon Bilis (2000), Typhoon Saomai, Typhoon Saomai (2000), Zhejiang

Title

Ruilong Li, Ph.D.

- Overview of typhoon risk in China - Hazard module - Vulnerability module - Modeled losses

Launched Northwest Pacific Basinwide Typhoon Model, including an extensive Separate wind and flood update for China and the losses are available in addition of South Korea Touchstone

2007 2010 2013 2014 2015

Released the first Introduced an Index Value Typhoon model for for our Industry Exposure China in collaboration Database to accommodate with the Shanghai the rapid development of Typhoon Institute China’s economy

Every year: • 30 form • 20 reach typhoon status • 5 reach super typhoon status

• Coastline ~14,500 km • Coastal mountains Beijing – Goulou Mountains Yellow River

– Wuyi Mountains

– Xandang Mountains • Major rivers Shanghai

– Yellow River ~6,000 km Wenzhou – Yangtze River ~ 5,500 km Yangtze River Fuzhou – Pearl River ~4,000 km • Coastal cities

– Shanghai Guangzhou – Wenzhou Pearl River – Fuzhou

– Guangzhou

Event Year Affected Provinces Econ Loss (CNY Affected Provinces Billions) 2004-2006 Rananim 2004 Zhejiang, Fujian, 22-30 Shanghai, Jiangsu, Anhui, Hubei, Jiangxi, Henan, Hunan Matsa 2005 Zhejiang, Shanghai, 16-24 Jiangsu, Shandong,

Liaoning Liaoning, Hebei, Tianjin

HebeiHebei Khanun 2005 Zhejiang, Shanghai, 13-17 Jiangsu

ShandongShandong Longwang 2005 Fujian 7-11 Bilis 2006 Fujian,Hunan,Guang- 22-30 Henan Damrey dong,Jiangxi,Guangxi, JiangsuJiangsu Anhui Shanghai Hubei Zhejiang Chongqing Zhejiang Rananim Saomai 2006 Zhejiang, Fujian, 10-15 Hunan Jiangxi Saomai Jiangxi Guizhou Bilis Fujian Fitow Damrey 2012 Jiangsu, Zhejiang, 3-8 Shandong, Hebei, Guangxi Guandong Liaoning Usagi Usagi 2013 Zhejiang, Fujian, 20-30

Hainan Guangdong, Guangxi Fitow 2013 Zhejiang, Jiangsu, 60-65 Fujian, Anhui, Jiangxi

- Understanding unique factors are key to modeling typhoons in China • Inland extent of precipitation footprint • Interaction of typhoons with inland weather systems • Effect of coastal mountains on typhoon precipitation • Effect of South China Sea monsoons

• Genesis point chosen from historical data • Track built using statistical and probabilistic information based on historical information • Wind characteristics added, such as central pressure and radius of maximum winds • Precipitation characteristics added, such as maximum rate and rainfall radius

low high

Weak (Cat 0-2) 0.5 Modeled 0.4 observed 0.3

Shanghai 0.2 0.1

0 Zhejiang Hainan Guangdong Fujian Zhejiang Shanghai Strong (Cat 3-5) 0.05 Fujian Modeled 0.04 observed

0.03 Guangdong 0.02

Hainan 0.01 0 Hainan Guangdong Fujian Zhejiang Shanghai

U.S. Gulf Coast China Coast

Winds(mph) at 0, 6, 12 h Winds(mph) at 0, 6, 12 h

Second-generation typhoon precipitation model introduced in 2007 – Accounted for different morphological types (circular  comma) – Accounted for terrain / coastal / weather / climate effect

2000 2007 2010

First-generation typhoon Third-generation typhoon precipitation precipitation model model introduced in 2010 introduced in NWP Model in – Improved transition between 2000 morphological types – Circular precipitation shield – Improved extratropical transition – No accounting for coastal / terrain – Improved coastal / terrain / weather / climate effects enhancements – Improvements based on the availability of additional data

• Bilis made landfall with winds of only 70 mph, but generated in excess of 600 mm of total precipitation • After landfall, Bilis turned southwest and affected 6 provinces over a period of 5 days • Bilis’ wind speeds were half those of Saomai, but the storm generated twice the amount of economic loss

Return Period Precip - Shanghai 800 50 Year Return Period Precipitation (mm) Modelmodel Historicalhistorical 600

400

Precip Amount Precip 200 Shanghai

0 50% 4% 2% 1% 0.40% Exceedance Probabilities Return Period Precip - Guangzhou 800 Modelmodel Historicalhistorical 600

400 Guangzhou Precip Amount Precip 200

0 50% 4% 2% 1% 0.40% Exceedance Probabilities

- Separate vulnerability functions are developed for wind and flood - Vulnerability assessment is based upon building attributes: – Occupancy: single-family, apartment, commercial, industrial, and CAR / EAR – Construction: masonry, steel, and reinforced concrete, etc… – Height: low (1 to 3 stories), medium (4 to 7), and high (8+) - Separate vulnerability functions are developed for building, contents, and time element coverages - Vulnerability functions account for duration of wind and rainfall - Other factors include building type, construction practice, construction material, and workmanship in the region

Single-Family Apartment Commercial

Non-Engineered Structure Marginally Engineered Structure Engineered Structure

• Poor workmanship in • Reasonably good • Well-designed roof and wall construction of roof and wall construction of roof and wall building elements • Weak foundation • Good foundation • Solid elevated foundation

Masonry Steel Reinforced Concrete

1.0 0.6 0.5 • Weak connections between • Strong frame structure • Strong frame structure building elements • Surface corrosion, rust- • Cracking and rebar • Pervious induced expansion expansion

High-rise (8+ stories)

Mid-rise (4 to 7 stories) Low-rise (1 to 3 stories)

• Marginally engineered structure • Engineered structure • Engineered structure • Highest damage ratio • Medium damage ratio • Lowest damage ratio

• Time-dependent replacement cost

• Time-dependent vulnerability

R(t) V(t)

CHINA

Wanda ’56 Rananim ’04

Winnie ’97 Insured Loss Loss Insured

AAL 10% 4% 2% 1% 0.4% 0.2% Exceedance Probability

TOT Shanghai TOT CHINA WND WND FLD FLD